Evaluation of the State of the Practice: Effects of Jointed/Discontinuous
Rock on Drilled Shaft Design
By
Karl A. Higgins, III, P.E.
Missouri University of Science and Technology (MS&T)
Course GeE 477 – Discontinuous Rock
Fall Semester 2011
ABSTRACT
Title: Evaluation of the State of the Practice: Effects of Jointed Rock on Drilled Shaft
Design
Author: Karl A. Higgins, III, P.E., MS&T Graduate Student
The author will examine the current state of the practice for evaluating drilled shaft
capacity for shafts socketed into rock of varying quality (from severely jointed and
weathered poor quality rock to unweathered, widely jointed, good quality rock).
Characterization of rock using varying methods of unconfined compressive strength,
Rock Quality Designation (RQD), Rock Mass Rating (RMR) and Geological Strength
Index (GSI) will be examined in the context of drilled shaft capacities. The author will
compare and contrast how drilled shafts bearing in soil or highly weathered rock versus
a rock socket behave. The author will compare the current practice state methods with
actual load tests of drilled shafts bearing in rock of varying quality for a recent
engineering mega-project, the $3 billion Northern Virginia High Occupancy Toll (HOT)
Lanes project currently under construction that resulted in the replacement of 40 bridges
many supported on drilled shafts.
Table of Contents
1.0 Introduction .............................................................................................................. 1
2.0 Description of Drilled Shafts ..................................................................................... 1
3.0 The Geomaterials Drilled Shaft Bear In .................................................................... 2
4.0 How Drilled Shaft Create Capacity ........................................................................... 3
5.0 Shaft Friction in Rock Sockets.................................................................................. 6
6.0 End Bearing in Rock Sockets ................................................................................... 8
7.0 State of the Practice ............................................................................................... 11
8.0 Case Study ............................................................................................................. 12
8.1 Test Shaft in IGM ................................................................................................ 13
8.2 Test Shaft in Rock .............................................................................................. 15
9.0 Closing ................................................................................................................... 17
1.0 Introduction
Engineers use drilled shafts as deep foundations to support a variety of civil
infrastructure projects and buildings. Drilled shafts are robust deep foundation elements
capable of carrying high axial compression, uplift, and lateral loads and thus are ideal
for certain unusual loading conditions such as traffic breaking, wind and seismic forces
on bridges.
At times, the loads required of drilled shafts are very significant, and engineers seek to
found the base of the drilled shafts into rock where capacities are generally higher. The
challenge with this objective is that relatively competent rock (rock that is not severely
jointed or weathered) may not be present at reasonable depths for drilling and
constructability purposes. As such, engineers are often required to form the base of the
drilled shafts into closely jointed and weathered rock.
The objectives of this paper are to examine the geologic factors that affect drilled shaft
capacity, compare and contrast the differences in shaft behavior for shafts bearing in
soil/weathered rock and rock.
The author will compare the current practice state
methods with actual load tests of drilled shafts bearing in rock of varying quality for a
recent engineering mega-project, the $3 billion Northern Virginia High Occupancy Toll
(HOT) Lanes project currently under construction that resulted in the replacement of 40
bridges many supported on drilled shafts.
2.0 Description of Drilled Shafts
Drilled Shafts are broadly described as a cast-in-place deep foundation whereby the
shaft is stabilized to permit the installation of reinforcing steel and concrete. Drilled
Piers are synonymous with Drilled Shafts, but “Caissons” are not.
Page 1
Drilled shafts
generally have larger diameters as compared to drilled, continuous flight, hollow stem
auger piles or “Augercast Piles,” but this distinction is changing as maximum Augercast
pile diameters are upwards of 48 inches. Drilled shafts are perhaps the most robust of
deep foundation types, and they are used for bridges, large buildings and major Civil
infrastructure projects due to their unique ability to resist high axial and lateral loads.
The design capacity of drilled shafts can vary between 30 to 6,000+ tons with diameters
from 30 to 120+ inches. The only limit on drilled shaft diameter is equipment capability.
Typical drilled shafts lengths are 20 to 90+ feet; however, exceptions as deep as 200
feet+.
3.0 The Geomaterials Drilled Shaft Bear In
Shaft bear in what engineers categorize as three material types: 1. Soil, 2. Intermediate
GeoMaterials (IGM), and 3. Rock.
Soil is further broken down into two types,
cohesionless and cohesive. Cohesionless materials include Sands and Gravels, and
non-plastic Silts that are deposited or weathered in place (i.e., residuum). Cohesive
materials are clays and sandy/gravelly clays with undrained shear strengths less than
5,000 psf and Liquid Limits greater than 50.
The geologic weathering process can change rock into soil. This transition between soil
and rock is generally vertical in the geologic lithology, with the degree of weathering
decreasing with depth. As the earth’s materials transition from rock to soil, engineer’s
have characterized the transitional materials as Intermediate GeoMaterials (IGM) or
weathered rock. IGM is stronger than soil and weaker than rock and posses both soil
and rocklike properties. IGM is subdivided into Cohesionless and Cohesive groups.
Cohesive IGM is defined as materials that exhibit unconfined compressive strengths in
Page 2
the range of 10 ksf < qu < 100 ksf. Cohesionless IGM is defined as very dense granular
geomaterials with SPT N60 values between 50 and 100 blows per foot.
Lastly, rock, a material with no consensus definition in geology or geotechnics, is
defined by engineers in drilled shaft design as a cohesive, cemented geomaterial that
can be identified on the basis of geologic origin. This definition of rock was provided by
the National Highway Institute (NHI) in their publication Drilled Shafts: Construction
Procedures and LRFD Design Methods dated May 2010. In the author’s opinion, this
NHI definition of rock is too vague.
4.0 How Drilled Shaft Create Capacity
For this discussion, the author will subdivide the explanation into two categories: shafts
bearing in soil/IGM and shafts bearing in rock. Shafts that bear in these two groups of
materials behave fundamentally different.
For shafts that bear in soil/IGM, engineers sum the side shear resistance based on
differing soil/IGM layers and then add the end bearing to determine capacity. Two
important equations are Rtot = ΣRSN + RBN and Rtot > F.S. X QTN.
provides a schematic of this equation.
Page 3
Figure 1 below
Figure 1
For cohesionless materials, engineers calculate the side shear resistance (shaft to
soil/IGM) by determining overburden pressures, converting these pressures into
horizontal pressures via earth pressure coefficients, and then determining the shearing
resistance between soil and the shaft concrete by estimating frictional interface
coefficients. Side shear in cohesive soils are estimated by determining the undrained
shear strength of the clay, determining the appropriate alpha factor (a reduction
coefficient), then multiplying the shear strength and the alpha factor by the
circumferential area of the shaft. End bearing for cohesionless or cohesive soils is
calculated conventionally per Terzaghi’s original theories on bearing capacity modified
for deep foundation effects.
Shafts with bases founded in soil/IGM require “movement” to mobilize side shear and
end bearing support. The shafts first engage side shear resistance, as relatively small
amounts of downward shaft movement are required to mobilize these resistances.
Once the side shear is fully mobilized, the end bearing becomes significantly engaged,
however, considerably more movement is required to mobilize full end bearing
resistance. At times, the amount of movement required to fully mobilize end bearing
(i.e., bearing shear) is excessive and cannot be tolerated structurally; hence, total
Page 4
capacity is not a function of the full end bearing resistance. Figure 2 below explains the
basic relationship between shaft response to top loading.
Maximum side resistance occurs
Maximum base resistance
at relatively small displacement
occurs at relatively large
and is
displacement
independent of shaft diameter
Figure 2
Drilled shafts with bases formed in rock sockets behave quite differently than described
for soil/IGM above. The primary differences are that the overburden pressures do not
affect the side shear of the concrete in the rock socket, and that axial support from the
soil overburden above the rock socket is ignored. Further differences include ultimate
capacity of the shaft being either from side shear in the socket, or end bearing, but not
the sum of both.
So why do we not sum end bearing and side shear from the socket for capacity for rock
supported shafts? The answer lies in the differences between soil and rock. Side shear
in the socket is developed by engaging the rock’s asperities (or the roughness of the
socket) with the cast-in-place concrete. Once the load is increased and the asperities
are sheared, the load is transmitted to the base. Unlike soil or IGM, that may remold
and re-adhere around the concrete after significant movement, rock does not behave
this way. The volume/shape of the socket is not appreciably affected by horizontal
overburden pressures and thus the “residual” socket side shear cannot re-engage.
Further, the asperities are sheared and thus if still resistive, are lower than the peak
strength.
Page 5
Why are the frictional side shear resistances of the soil overburden (residual or
deposited soil) and IGM above a rock socket ignored?
The answer lays in strain
compatibility differences between soil and rock. As explained above, relatively small
amounts of downward shaft movement are sufficient to engage and then mobilize fully
side shear resistance in soil/IGM. To mobilize the side shear in a rock socket, even
smaller amounts of movement are required. Since the shaft is concrete and stiffer
(much higher modulus) than the soil overburden/IGM, as load is applied to the top of the
shaft, the shaft responds by load shedding top down.
The less stiff upper soils
ultimately yield and transfer stress to the stiffer materials comprising the rock socket.
This load shedding characteristics of rock socket will be demonstrated in the case
history portion of the paper.
For these reason, engineers either design shafts bearing in rock sockets as side shear
or end bearing only, but not a combination of the two. Further, the side shear from
soil/IGM above the socket is ignored from a capacity perspective due to load shedding.
As explained, shafts bearing in rock sockets behave fundamentally different than for
shafts bearing in soil/IGM.
5.0 Shaft Friction in Rock Sockets
Drilled shafts are large, highly loaded elements that are difficult (and expensive) to load
test. As such, there is not frequent field testing of drilled shafts to confirm design
assumptions. What engineers are ultimately interested in is the frictional resistance
between the concrete and the shaft rock socket.
The engineering design process
results in some basic assumptions or characterizations: roughness coefficient, jointing
of the rock, and soil materials of in filled joints. The drilling process itself can affect rock
socket roughness and obviously the rock type is a factor. Cleanliness of shaft sockets
Page 6
prior to concrete pouring is also a concern.
Despite all these challenges, some
researchers have presented theories and empirical relationships for other more easily
measured properties to estimate shaft socket roughness.
Two of the more common procedures for estimating socket roughness are from O’Neill
and Reese’s research and from Horvath and Kenney’s research. These authors choose
different, more easily measured rock properties to indirectly estimate socket roughness.
O’Neill and Reese suggest socket roughness is a function of unconfined compressive
strength, qu, and a term called alpha, a, which in turn is a function of rock jointing
(measured by the RQD process). The more severely jointed the rock is, the lower the
RQD and alpha values are respectively. The weaker the rock is, the lower the qu value.
Since alpha and qu are directly proportional in the equation, the weaker and more
jointed the rock is, the lower the socket friction.
Since RQD and qu are common
engineering measurements on boring/coring logs for cored rock, this correlation is
significant in engineering practice.
O’Neill and Reese Equation for
socket shaft resistance. Pa =
atmospheric pressure, qu =
unconfined compressive
strength of the rock, aE is the
reduction factor.
Horvath and Kenney suggested socket friction, fSN = 2.5(qu)0.5, is related only to the
rock’s unconfined compressive strength, qu, also an engineer property that is often
measured in the lab and reported on boring/coring log reports.
There are undoubtedly more research and correlations related to estimating socket
shaft resistance in rock. What is interesting about the above author’s work is that
neither introduce the basic rock type into the equation. One can imagine that a finegrained siltstone (sedimentary rock), that is not significantly jointed and has reasonably
high qu value (say 4,000 psi) would have a lower socket roughness than a coarse
Page 7
grained granite (igneous rock) with similar properties of jointing and strength. However,
neither equation above takes rock mineralogy into consideration.
6.0 End Bearing in Rock Sockets
Like shaft resistance, there are several notable authors who have conducted research
into base support for drilled shafts with rock sockets. Prakoso and Kulhaway used a
series of load tests for shafts bearing in rock to determine a relatively simple empirical
equation qBN = Nc* qu.
Nc=2.5 is
recommended
when qu alone is
used to
Fig. 3
characterize
rock mass, Nc is
not a F.S.
Base diameter does not appear to affect the equation (different than for soil/IGM
supported shafts) and again rock type is not part of the equation. For Kulhaway’s
equation to be valid, the rock must either be massive or tightly jointed to a depth of at
least 1D below base, and the base must be clean base and visually verified.
When data are available on the spacing and condition of discontinuities in rock beneath
the base, the method described in the Canadian Foundation Engineering Manual
(Canadian Geotechnical Society, 1995) provides a more refined estimate of N c
Page 8
(Nc=3Kspd) but uses the same equation qBN = (3Kspd) qu. For this method to be valid,
discontinuity spacing is at least 1 ft, and discontinuity aperture does not exceed ¼ inch.
Canadian
Geotechnical
Society Method relies on
knowledge of spacing an d
discontinuities below shaft
base.
The NHI/Federal Highway Administration’s recent publication addresses end bearing in
yet another equation, this time relying on the engineering geology properties of Rock
Mass Rating (RMR) and qu to determine end bearing.
NHI/FHWA Method for base
resistance in rock
Lastly, Hoek and Brown have developed their own methodology for determining base
resistance, but this time it relies on the Geologic Strength Index (GSI) as the socket
Page 9
characterization tool.
Hoek believes that GSI is a better method for determining
engineering properties of rock than is the RMR method. According to Hoek, GSI is
based upon an assessment of the lithology, structure and condition of discontinuity
surfaces in the rock mass. What is appealing about Hoek’s ideas is that for the first
time, the roughness of the rock mass (a characteristic of mineralogy) and weathering
are factors. The basic premise of Hoek’s GSI theory is shown graphically below in
Figure 4.
Decreasing Surface Quality
Decreasing
interlocking
of rock
pieces
10090
10-0
Fig. 4
A GSI of 100-90 would describe
a massive, widely jointed rock
with rough, fresh surfaces
whereas a GSI of 10-0 would
describe a rock that had
indistinguishable blockiness and
weathered, slickensided
surfaces.
Hoek’s equations for end bearing:
S, a and mb are
Hoek strength
parameters
based on GSI
Page 10
7.0 State of the Practice
The National Highway Institute (NHI) and the Federal Highway Administration (FHWA)
have long been at the forefront of geotechnical research and publication. The two
organizations have combined and prepared many geotechnical, practice based design
manuals intended to help ensure the nation’s infrastructure is designed adequately. A
recent, 2010 NHI/FHWA Publication titled Drilled Shafts: Construction Procedures and
LRFD Design Methods does a good job in synthesizing the cumulative research and
design approaches for drilled shafts. One of the appealing items of this publication is
that it presents several author’s perspectives allowing the practitioner to determine
which theories to apply. Some of the more commonly referenced authors are O’Neill
and Reese (combined research) and Kulhaway.
The manual provides specific
equations and reference materials for shafts bearing in differing materials.
After conducting research, Prof. Reese (University of Texas) went on to form a software
company called Ensoft Inc. The design procedures captured in the FHWA manual(s)
were converted into software for geotechnical engineering analysis purposes.
One
notable software for analyzing the axial behavior of shafts is Ensoft’s Shaft® program.
The Shaft program is capable of analyzing shafts bearing in a variety of soil and IGM
materials, and rock of differing type and weathering. Pertinent input parameters for
shafts bearing in rock include rock unconfined compressive strength (qu), Young’s
Modulus, RQD, spacing and thickness of the discontinuities.
Kulhaway’s method for shafts bearing in rock sockets is utilized by Shaft® and is
summarized below.
1. Calculate required socket length based on side friction alone
2. Compute settlement of shaft by adding elastic shortening (PL/AE) of the shaft
itself to the amount of settlement required to mobilize end bearing, q p assuming
the full load of the shaft is taken by end bearing
Page 11
3. If computed settlement < 0.4 inches, side resistance will dominate and little end
bearing will be engaged
4. If computed settlement > 0.4 inches, bond in socket will be broken and end
bearing will be engaged
The concept is that side resistance and end bearing will not be developed
simultaneously, and that Qult ≠ Qside friction + Qend bearing (different than IGM shafts), rather
Qult = Qside friction or Qend bearing depending on settlement amount.
While Kulhaway’s methods are cited for “rock” (which is presumably hard rock as the
software manual describes), there are alternative methods for weak rock as presented
by O’Neill.
The software manual uses the term IGM (previously described)
synonymously with weak rock. Input parameters for weak rock/IGM include qu, RQD,
core recovery percentages, Young’s modulus, a description of the joints (open or
closed), and whether or not the socket would “smear” during drilling.
In summary, one can use the FHWA manual and perform hand calculations or use
Ensoft’s Shaft® program which is based on this manual, both of which appear to be the
most progressive state of the practice for drilled shaft design.
8.0 Case Study
The author was fortunate enough to be one of several principal geotechnical engineers
involved with the design and construction of the I-495 High Occupancy Toll (HOT) lanes
project, a Civil Engineering mega project.
This $3billion project included the first
modernization of the DC Capitol Beltway since its initial construction, and resulted in the
replacement or addition of more than 40 bridges.
Many of the bridge piers were
supported by Drilled Shafts bearing in IGM and Rock.
Because of the size and
complexity of the project, there was sufficient money to do several sacrificial test shafts
using an Osterberg Load Cell (or O-Cell). The author reviewed the test data and there
Page 12
are examples that support the theories on the differences between shafts bearing in
IGM and rock presented in this paper.
8.1 Test Shaft in IGM
As previously mentioned, shafts with bases founded in IGM require “movement” to
mobilize side shear and end bearing support.
The shafts first engage side shear
resistance, as relatively small amounts of downward shaft movement are required to
mobilize these resistances. Once the side shear is fully mobilized, the end bearing
becomes significantly engaged, however, considerably more movement is required to
mobilize full end bearing resistance. The graphic in Figure 5 below depicts the load
settlement graphs of a test shaft with the base bearing in IGM. Note the significant
amount of movement on both sides of the O-cell which represent the development of
shaft friction and base resistances.
Page 13
Fig. 5
Note significant base of shaft movement in
IGM during load test
Page 14
Table A above is the corresponding summary table showing the side shear resistance
developed in the shaft during the test. Note that there is significant side shear (1.26 to
3.45 ksf) in the residual soil above the O-cell, and even higher shaft resistances at and
below the O-cell where denser residual soil and IGM are present. Further, the base of
the shaft was in the process of shear engagement, but did not likely experience full
base resistance shear mobilization (typically thought to be 10% of the shaft diameter,
which would be over 5 inches in this example). In summary, the O-cell test data fits
nicely with the theories of shafts bearing in IGM whereby first side shear resistance is
mobilized, then ultimately the base of the shaft, and that both side friction and end
bearing are added to determine shaft capacity.
8.2 Test Shaft in Rock
Side shear in the rock socket is developed by engaging the rock’s asperities (or the
roughness of the socket) with the cast concrete. Once the load is increased and the
asperities are sheared, the load is transmitted to the base. Unlike soil or IGM, that may
remold and re-adhere around the concrete after significant movement, rock does not
behave this way.
The volume/shape of the socket is not affected by horizontal
overburden pressures and thus the “residual” socket side shear cannot re-engage.
Further, the asperities are sheared and thus if still resistive, are lower than the peak
strength.
To mobilize the side shear in a rock socket, even smaller amounts of movement are
required. Since the shaft is concrete and stiffer (much higher modulus) than the soil
overburden/IGM, as load is applied to the top of the shaft, the shaft responds by load
shedding top down. The less stiff upper soils ultimately yield and transfer stress to the
stiffer rock socket. This load shedding characteristics of rock socket is demonstrated in
the load test below (Figure 6).
Page 15
Fig. 6
No upward movement of Shaft
Note small base of shaft movement in Rock during load test
(<0.10 inch). Side shear in socket is dominating capacity.
Page 16
Table A above is the corresponding summary table showing the side shear resistance
developed in the shaft during the test. Note that this time, there is no mobilization of
side shear in the residual soil above the O-Cell (there is one segment (strain gauge
level 1-2) where 2.7 ksf of shear is mobilized, but this location is very near the top of the
socket in more competent IGM. Since the settlement amounts of the socket are small
(less than 0.10 inch), side shear in the socket is dominating resistance, and in fact end
bearing is not likely even engaged. There is insufficient shaft movement upward to
engage side shear in the residual soil above the socket (hence why it is ignored). This
shaft behavior corresponds very well with the theories on shaft loading in sockets
presented in the FHWA manual.
9.0 Closing
Shafts bearing in discontinuous rock (IGM and highly jointed rock) behave
fundamentally different than shafts bearing in rock sockets. One must appreciate these
behaviors when designing shafts for support. It would be a mistake, for example, to
include side shear from the residual soils above the socket to a shaft’s capacity. The
current state of the practice is adequately captured in NHI/FHWA’s 2010 publication on
the design of drilled shafts.
C:\UM S&T\GE 477 Discontinuous Rock\Assignments\Term Project 1\GeE 477 Term Project 1 Higgins Paper.doc
Page 17
References
(1) “Drilled Shafts: Construction Procedures and LRFD Design Methods,” NHI
Course 132014, FHWA-NHI-10-016, FHWA GEC 010, U.S. Dept. of
Transportation, Federal Highway Administration, May 2010
(2) “Computer Program Shaft Version 6.0”, by Lymon Reese, et al, 2007
(3) “Analysis and Design of Drilled Shaft Foundations Socketed into Rock”,
prepared by Cornell University for the Electric Power Research Institute, EPRI
EL-5918, August 1988, by Authors Carter and Kulhaway
(4) Power point presentation, “Determination of Unit Tip Resistance for Drilled
Shafts in Fractured Rock using the Global Rock Mass Strength”, by Truzman,
Corley and Lipka (undated)
(5) “Improving Foundation Design in Rock: Load Test at Burma Road Overpass”,
FHWA-WY-09/10F, University of Wyoming, December 2009, Author John
Turner, Ph.D.